"contour" of a likelihood

Dear Experts,
I would like to know whether the member function “contour” of the classe "RooMinuit"
performs a likelihood profiling or it is simply a likelihood scan ?

Example of code:
RooAbsReal* NLL;
NLL = (*TotalPDF)->createNLL(*dataSet,Extended(true));
RooMinuit RooMin(NLL);
myFrameNLL = RooMin.contour(*GetVar(*TotalPDF,“AfbS”),*GetVar(*TotalPDF,“FlS”),1.0,2.0,3.0);

In other words, does the “contour” re-minimise the likelihood with respect to the nuisance
parameters at every step of the parameters “AfbS” and “FlS” ?

Many thanks.

  • Mauro.


The contour is obtained by profiling all the other parameters, i.e. by doing a minimisation in N-2 dimension.